TY - JOUR
T1 - H∞ Optimal One-Step-Ahead Output Feedback Control of Discrete-Time Systems
AU - Yaesh, I.
AU - Shaked, U.
PY - 1992/8
Y1 - 1992/8
N2 - The discrete-time, one-step-ahead, standard four-blocks H∞ optimal control problem is considered. The output feedback problem is transformed into a one-step-ahead H∞ optimal state-estimation problem, and the existing solution to this problem is then used to derive the controller. It is assumed that only previous measurements are available for control. Two alternative results are obtained. The first, which has an apparent merit over the existing methods, requires the solutions of two coupled Riccati equations, where the necessary and sufficient condition for the existence of the H∞ optimal result requires only the positive semidefiniteness of these solutions. The second result applies two decoupled Riccati equations with a coupling constraint, as in the continuous-time case, and two additional conditions. The derivation procedure is most simple and may thus provide considerable insight into the solution of the discrete-time standard problem. A similar procedure can also be used to obtain a solution to problems where the current measurement is available.
AB - The discrete-time, one-step-ahead, standard four-blocks H∞ optimal control problem is considered. The output feedback problem is transformed into a one-step-ahead H∞ optimal state-estimation problem, and the existing solution to this problem is then used to derive the controller. It is assumed that only previous measurements are available for control. Two alternative results are obtained. The first, which has an apparent merit over the existing methods, requires the solutions of two coupled Riccati equations, where the necessary and sufficient condition for the existence of the H∞ optimal result requires only the positive semidefiniteness of these solutions. The second result applies two decoupled Riccati equations with a coupling constraint, as in the continuous-time case, and two additional conditions. The derivation procedure is most simple and may thus provide considerable insight into the solution of the discrete-time standard problem. A similar procedure can also be used to obtain a solution to problems where the current measurement is available.
UR - http://www.scopus.com/inward/record.url?scp=0026902124&partnerID=8YFLogxK
U2 - 10.1109/9.151118
DO - 10.1109/9.151118
M3 - ???researchoutput.researchoutputtypes.contributiontojournal.article???
AN - SCOPUS:0026902124
SN - 0018-9286
VL - 37
SP - 1245
EP - 1250
JO - IEEE Transactions on Automatic Control
JF - IEEE Transactions on Automatic Control
IS - 8
ER -